Mashrura Musharraf Investigates Methods to Improve Safety Training for Offshore Oil and Gas Workers
by Wade Kearley for OceansAdvance Inc.
An alarm blares and loud voices rush past your room. Donning a survival suit, you open the door. Visibility is poor in the emergency light. As one person among dozens on an offshore platform, you know that every decision is critical for safety. Your heart races and your breathing quickens as you follow the egress route. You quell the urge to run, even as a co-worker rushes by. At the exit, he’s yanking on Stairwell 1 door. It’s jammed. So you lead the panicked worker up Stairwell 2 to join others at the muster station. Light flashes on the screen: “Congratulations. You have safely completed the test.” You let go of the gamepad and, with a sigh, lean back in your chair.
Dozens of trainees have navigated this same virtual environment (VE) in a computer laboratory in the engineering building at Memorial University’s St. John’s Campus. These tests are run by Mashrura Musharraf, an award-winning computer engineer, and Ph.D. candidate. She’s agreed to meet and talk about her work and her background. We are in the boardroom of the Ocean Engineering Research Centre where she invites me to sit at a long table. In her late twenties, she’s dressed casually, and though she sits upright, she appears relaxed and curious.
The Future for Emergency Training
For Musharraf, this type of VE training is the future for emergency preparedness and safety training for offshore workers. She and her team designed the all-hands virtual emergency response trainer (AVERT) to enhance offshore emergency response training. The VE simulates a detailed model of an offshore oil installation platform through which participants navigate using a gamepad.
In her study, they create emergency scenarios by introducing a range of activities, from muster drills to more complex emergency evacuations. A key to this innovative research project, says Musharraf, is the database of human responses in simulated offshore emergencies that they are compiling. This information will aid in the development of training tools for offshore workers. Musharraf says this training can have an immediate positive impact offshore by helping to improve human reliability. “We say that people are reliable who have a higher probability of successfully completing a task.”
Her research project was conceived by Dr. Brian Veitch, Memorial’s Industrial Research Chair in Safety at Sea. Speaking by telephone from his office, he’s supportive of Musharraf’s contributions. From his earlier work, he saw a benefit to using simulators to replicate virtual marine environments. “I wanted to use virtual environments to do experiments with humans to analyze human reliability,” recalls Veitch. “So I recruited Mashrura in 2011 and she did it. Her work was quite innovative.”
The Path of Certainty
Musharraf grew up in Rangpur, a small city northwest of Dhaka, the national capital. Her father was the regional manager for a pharmaceuticals company, and her mother was the deputy general manager of a bank. At school Musharraf’s fondness for solving problems led her to excel at one subject in particular. “As a kid I liked math,” she says. “Your answer is either right or it is wrong. There is no grey area. I want to know things and I want to eliminate uncertainty as much as possible.” Her final grades were good enough to get her accepted into Bangladesh University of Engineering and Technology (BUET), Dhaka—provided she passed the entrance exam.
“The test to be admitted into university is very hard,” says Musharraf, “and when you graduate from high school you need additional coaching to be competitive. So a group of us moved to the city and enrolled in a coaching school for three months.” The work was intense. Every day brought challenges associated with living in residence. “I had never traveled before. I had to leave my parents. I had never lived in a hostel. This is not what you are trained for when you leave home.”
One day, about two months into the program, feeling stressed by changes in her life, she became tearful on the phone with her mother. “I was thinking that she will comfort me. But she told me, ‘Look at the bigger picture. Do you want to be an engineer?’ I told her ‘Yes,’ and she said, ‘So suck it up. Do what you went there to do. I know my girl. She is strong. You can do this. Do not make decisions on little things… Focus on the bigger picture.” Musharraf laughs at the recollection. “I was really mad at her that day, but now I can be grateful because she’s probably the reason I didn’t quit.”
Looking for the Next Challenge
As predicted, Musharraf passed the entrance exams and graduated with a bachelor’s degree in computer science and engineering, discovering, along with the way, an interest in artificial intelligence and data mining. After graduation she signed on with a software engineering company, Dohatec, building websites for private and public companies.
“It was exciting because it was my first job where I could apply my skills,” she says. She learned process documentation, working to deadlines, and acceptance for one’s team role. But the best part for her was seeing people use what she built. “For the first time, it felt like I was doing something to make a difference.”
But after about 18 months the thrill of repetitive teamwork faded. “I like working in a team, but I need to be challenged,” she says. “I was pretty young, but I knew that was not something I wanted to do for the rest of my life.” And she knew, from her academic experience, that she preferred research. “You are always solving new problems,” she says, so she enrolled in her master’s degree at BUET.
However, for a variety of reasons she left that master’s program, and as a newly-wed with the blessing of her and her husband’s families, began contacting international professors with similar research interests. One of those was Dr. Syed Imtiaz, a professor in process engineering at Memorial in a place called Newfoundland and Labrador. A few days after sending a CV, she heard back. He’d forwarded her resume to Dr. Faisal Khan who showed it to his colleague Dr. Veitch. “Because of my computer engineering background, he wanted to hire me for an upcoming project related to artificial intelligence.”
In less than two months she boarded a plane for the 38-hour flight to Canada by way of Doha and London, arriving at St. John’s airport in September of 2011 when the weather, “was not too much of an adjustment.”
The Uncertainty of Human Behaviour
A big challenge for Musharraf, as she worked towards her master’s, was the uncertainty of the data she was using to assess human reliability. Unlike the black and white mathematics of her school days, she was confronted with large areas of grey. “I defined the data I needed but I began to realize that it is very difficult to get a reliable number. It is all very subjective,” she says. “I reviewed the literature and found that others using expert judgment to assess human reliability mentioned limitations such as inconsistency, conflict among experts, and bias.” She needed numbers that were independent of personal bias.
Veitch had anticipated this. “Human performance is difficult to assess in any quantifiable way and it is particularly difficult in emergency situations because you don’t have any data. People aren’t making notes during emergencies,” he says. “Using simulators where you can simulate the natural ecosystem and record responses allow data collection instead of reliance on opinions.”
Based on this assumption, Musharraf sought to generate objective data. Veitch supervised the process. “We did an experiment where we developed changing scenarios in the virtual environment that were controlled,” he says. “We had participants go through those and we collected data on their performance.”
Using that data, Musharraf applied the mathematical model to predict people’s reliability during emergencies. She also demonstrated how a Bayesian network could be used as a diagnostic tool to identify individuals’ strengths and weaknesses, which she says “can shape the path of adaptive training.”
Take What You Need and Leave the Rest
Based on the strength of that work, she completed her master’s degree and won a prestigious award in 2016 from the Research and Development Corporation to continue research for her Ph.D.
Musharraf and team are on the cutting edge of VEs that are redefining how human reliability is assessed and how workers can be trained to make the most of their capabilities. A goal of her current research is to generate machine intelligence to populate the VE with avatars that behave as real people might. “Advances in technology will allow the application of virtual environments and machine intelligence to create realistic scenarios and make the training that much more effective,” says Musharraf who will graduate with her Ph.D. in 2018.
“I will know my career has been a success if I can look back and say I enjoyed the journey,” she says. When she started her Ph.D. she’d heard a great deal of “scary stuff” about how nothing goes according to plan. But what she has discovered is that all that really matters is how good you are at solving the problems as they arise. That involves paying attention to criticism but not being bowled over by it.
“Most people who are criticizing you want your work to get better. So if you can get over the part about feeling bad, then you can use criticism and consider if there is something there that can help you. And if you have an open mind, you can usually find ways to move ahead,” she says. “But it is also important to have your own opinions. Just because someone says something is not good, that is not a reason to leave it.”